<< Previous Entry Next Entry >>
Journal Entry

Monday, September 13, 2004

Bacon-Erdös numbers

I had come up with a variant on this, and was amusing some people at Worldcon with the notion (and bewildering others -- but Cory Doctorow, at least, laughed his ass off). As usual, Google reveals I'm a couple years late to the party.

My proposal was to find someone who had both an Erdös number and a Kevin Bacon number, and use them to produce a conversion function. In one glorious stroke, a thousand lifelong dreams could thus be fulfilled -- everyone in Hollywood could have an Erdös number, and everyone at MIT could have a Kevin Bacon number.

As it turns out, the preliminary work has already been done. Brian Greene, for instance, has an Erdös number of 3, and a Bacon number of 2. Thus, my proposed conversion function (allowing edges in the unified Bacon-Erdös graph to represent two people either appearing together in a movie or coauthoring a paper) is as follows:

Finding: an actor with a Bacon number of N has, at most, a Baconized Erdös number of N+5. Similarly, an academic with an Erdös number of M has, at most, an Erdösinated Bacon number of M + 5.

(My initial lines of research, proposing to go through Dolph Lundgren or Natalie Portman, would surely have yielded much less powerful results.)

The emphasis of previous Bacon-Erdös research, however, has not been on unification, but rather on those individuals with authentic claims to both direct Erdös numbers, through actual academic coauthorship and to direct Bacon numbers through actual screen acting. Thus the canonical Bacon-Erdös number is the sum of an individual's separately earned Erdös number and Bacon number, and this --we learn -- is what is devoutly to be sought. The aforementioned Brian Greene and Dave Bayer are tied for the world-record lowest Bacon-Erdös sum of 5.

Strictly construed, my own Bacon and Erdös numbers are both infinite, since I have never coauthored a paper for a refereed academic journal, nor acted in a major Hollywood film. If we extend the definition of coauthorship, though -- allowing, say, coauthorship of unpublished novels to count toward Bacon numbers, and coauthorship of letters to the editor and the like to count toward Erdös number, we could be nice and award me an extended Bacon-Erdös sum of 6 -- going through my novel co-author David Ackert, who has a proper Bacon number of 2, and my Dad, who has a proper Erdös number of 2 (he's listed as D.M. Rosenbaum, co-author of Daniel Kleitman).

This is a bit iffy, however, I must admit.

If Ramin (aka David Ackert) were to co-author a scientific paper with my Dad (aka David Rosenbaum), however, Ramin would tie for the world record Bacon-Erdös sum. If they also acted in a film together, my Dad would also tie for the world record Bacon-Erdös sum -- which seems a reasonable quid pro quo.

I think you guys should get cracking.

Posted by benrosen at September 13, 2004 10:34 PM | Up to blog

Actually, if you read the book "Linked" by Albert-Laszlo Barabasi, he's done this work already.

Posted by: allaboutme at September 14, 2004 02:41 PM

Shoot. Late to the party as always. Thanks, allaboutme...

Posted by: Benjamin Rosenbaum at September 14, 2004 03:06 PM

Actually Dan Kleitman, a mathematician, has Bacon number 2 and Erdos number 1.


Posted by: Suresh at September 14, 2004 04:26 PM

Only if you count the less strict rule which construes a technical advisor as "being in" a movie. Although the Erdos Number Project site claims he appears briefly, I think they're kidding, as the gif they link to is a photoshopped gag.

I don't know about Bruce Resnick, who is reported to have a Bacon-Erdos sum of 3. Does being an extra count for canonical Bacon numbers, or do you have to have a credit in the film?

I think we can still argue that B<=E+5 for strict interpretations of "being in a movie with", and B<=E+3 for looser interpretations.

I believe I last saw Danny Kleitman at my Dad's 65th birthday, where (I believe) he told some funny stories intended to embarass my Dad -- I don't remember what they were exactly. He's a great guy.

Posted by: Benjamin Rosenbaum at September 14, 2004 04:43 PM

If this is true, and Dan Kleitman and Bruce Reznick both have Erdös/Bacon numbers of 3 -- and given that the lowest possible fresh Erdös number, since Paul Erdos is no longer with us, is 2 -- there is only own man who can break the world record: Kevin Bacon.

Posted by: Jim at September 14, 2004 04:50 PM

Slim Pickens, according to the research done for the book "Linked", is apparently the person you really want to find out what your "Erdos" number is. Forget Kevin Bacon. Small time.

In other words, the equivalent of an Erdos in the entertainment world is really Slim Pickens.


Posted by: allaboutme at September 14, 2004 09:25 PM

Dear Ben Rosenbaum

With all due respect, you may be up-and-coming, but you've reinvented the wheel, and wrongly at that. Dang, I should have skipped the college math courses I teach and just come to Noreascon.

The lowest sum of Bacon and Erdos Numbers is well-known to be Bruce Reznick.

As I said on Making Light
Open thread 23
May 26, 2004 11:24:26 PM EDT

"And I know the guy, Bruce Resnick, who has the lowest known total of Erdos Number + Kevin Bacon Number. He has an Erdos Number of 1 (coauthored with Erdos) and a Kevin Bacon number of 2 (appeared in "Goodwill Hunting" with someone who was in some other movie with Kevin Bacon). Unless Kevin Bacon writes a math paper..."

Then on Open thread 23
May 27, 2004 9:32:15 AM EDT


Fun pages about Erdos Numbers:


Some Famous People with Finite Erdös Numbers

Thrill at how "Felipe Voloch found what seems to be the oldest mathematicican known to have a finite Erdös number, Richard Dedekind (1831-1916). His number is at most 7, via this path: H. Weber -- W. Jacobsthal -- R. Fuchs -- L.Hopf -- A. Einstein -- E. Straus -- P. Erdös."

Laugh at how, among Nobel Laureates, Albert Einstein [1921, Physics] has an Erdos Number of 2, while Erwin Schrödinger [1933, Physics] has an Erdos Number of 8.

But wait... there's more!

Check out:
The Erdös Number Project Extended

to answer these burning questions:

* Who is The Youngest to Gain an Erdös Number?
* Which Movie Star has the lowest Erdös Number?
* Does Joan Baez have an Erdös Number of 3 ?
* Do any animals have an Erdös Number ? Is it possible
* Just what are the principles of Mathematical
Genealogy ?
* Did my Partner Acquire an Erdös Number on marriage ?
If Not, How can he/she acquire a finite one?
* Can a character in a novel gain an Erdös Number ?
* "Basic Rule for Assigning Erdös Numbers: Was enough
Chutzbah involved in the claim for it to be taken
seriously ???"

Then I started a blog for precisely what Ben Rosenbaum
stumbled onto later and in which he is now stumbling


Especially look at the initial thread, which is all about Erdos numbers IN SCIENCE FICTION by way of "Asimov Numbers":


Anyway, congratulations on writing science fiction, and thanks again for following my references here.

To see MY Erdos Number derivation, since it also connects with your Dad's coauthor Daniel Kleitman, via 3 Nobel laureates:



Jonathan Vos Post
(erdos Number = 5)
over 15,000,000 hits/year
#4 or so on Google and Yahoo for "science Fiction"

Posted by: Jonathan Vos Post at September 15, 2004 12:48 AM

Hope I didn't sound too snotty in previous posting. Maybe it stings to have my 15-year-old son, a Junior in college, who's already several times professionally published, planning to coauthor with a Graph Theory professor in order to have a lower Erdos Number than me, his Dad.

I went to Caltech with Bruce Resnick, but I'm sure that he doesn't need me to defend his honor. Saw him at a college reunion, too. I was feeling ebulient, maybe after too much champagne. I was boasting about my coauthoring with Ray Bradbury and Arthur C. Clarke, and so on, when someone nudged me. "See that guy over there? Doug? He was in the class a couple of years before you. HE won a Nobel prize for discovering superfluid helium-3 in his first year of grad school."

Boy, that put me in my place.

Maybe I should empathize with you. I often have Math papers rejected by referees who point out that some obscure guy in some remote country had the same result as me, sometimes years earlier.

Posted by: Jonathan Vos Post at September 15, 2004 12:56 AM

Wow, that's pretty extensive. Little did I know my modest inquiries into the Bacon-Erdos field were preceded by such broad and thorough investigations! I am honored, Dr. Von Post!

And bravo to Von Post fils for kicking academic ass. You must be proud (as well as jealous). He reminds me of my friend Terri, our high school circle's own Buckaroo Banzai, who was building plasma generators in her basement for her science project while we were doing "our friend the beaver" dioramas and the like.

Asimov numbers seem slightly less compelling, since collaboration is something of an anomaly in SF, while it's the rule in academic papers and movies. Plenty -- maybe most -- SF luminaries have never collaborated at all, and thus would have infinite Asimov numbers, yet they are clearly interconnected. Hmm.... yet a rule like "were in a magazine or anthology together" seems too loose...?

Posted by: Benjamin Rosenbaum at September 15, 2004 12:44 PM

Oh, by the way, Dr. VP, can you confirm that Bruce Resnick actually *appears* in GWH? Does he have a line in the film? Is he in the credits for an onscreen role? Then we would have B = E + 3 even under strict definitions.

Posted by: Benjamin Rosenbaum at September 15, 2004 12:46 PM

I don't know whether Bruce Resnick appeared as an extra in Good Will Hunting, but Daniel Kleitman did. Read Kleitman's own account at:


Thus, if appearing as an extra establishes an edge in the Bacon graph, then Kleitman indeed has a Bacon number of 2 and an Erdo"s number of 1.

Posted by: Aaron Archer at September 15, 2004 02:44 PM

It isn't true that only Kevin Bacon can break the record by getting a combined Erdos-Bacon number (CEBN) of 2. Any of the existing holders of Erdos number 1 could appear in a movie with Kevin Bacon and acquire a CEBN of 2.

Joe Ganley (Erdos number = 3, infinite Bacon number)

Posted by: Joe Ganley at September 16, 2004 04:26 PM

Dear Ben Rosenbaum, Aaron Archer, & Joe Ganley:

Erdos himself refers to "living" as meaning "doing mathematics" and "dead" meaning not doing mathematics, regardless of whether or not your body moves around.

Oddly, this means that one can be dead for many years, and suddenly live again.

If my son played guitar, he'd belong in Buckaroo Banzai's rock band. However, that's one of the few things my son does not do, although I was a rock journalist, publishing "Sound Options" in the mid 1970s on the Jersey Shore for 25,000 readers; and I wrote the lyrics for "Panic" which aired once as a punk song on MTV.

There is much truth in "collaboration is something of an anomaly in SF" with notable exceptions (Eando Binder, Pournelle & Niven & Barnes; The Strugatsky Brothers). But people who publish a lot only have to collaborate once in a long while to build the collaboration web.

Further, Isaac Asimov was also a professor of Biochemistry, and coauthored with a couple of people who only wrote Biochemistry, so Asimov links the web of Science Fiction authors with the web of 1,000,000+ biomedical paper authors.

If you check my blog, you'll see that there is a category of writer-directors who collaborated with Asimov in the sense of adapting his works to TV or film. Thus we can connect Asimov Number with Bacon Number.

Asimov himself was sad that nobody ever referenced his PhD dissertation in an academic paper. I promised him that I would. he may be dead, but I shall honor that promise.

As to "a rule like 'were in a magazine or anthology together' seems too loose...?" -- you are absolutely correct, although Robert Silverberg published an essay on such linkages.

A difference of opinion exists as to posthunmous collaborations. Erdos had many (papers started when he was alive, and finished later). Bear, Brin, and Benford were posthumous collaborators with Asimov, through their estate-authorized Foundation sequelae. But Hugo- and Nebula-winner Dr. Geoff Landis (whose day job is on the Mars explorer robots) insists that those are not true collaborations with Asimov.

Then there is the idea of weighting the links. If you coauthor 2 papers with Erdos, that counts extra. If you had 100 coauthors on one paper (common in particle physics) then your linkages are very lightweight indeed.

If you check out that first thread of my blog, you'll see that the whole subject is very serious -- for example, the network of who had sex with whom, in a world of HIV.

I heard some great stories from my buddy Allen Ginsberg abot this, but I dare not type them here and get general Ashcroft going medieval on your blog.

If you follow up with Cory Doctorow, do tell him that we settled our argument peacefully, and that you're still a funny guy.

Thank you for your rapid response. Hope to see you at the Worldcon in Glasgow!

Posted by: Jonathan Vos Post at September 16, 2004 06:55 PM
<< Previous Entry
To Index
Next Entry >>